CN115731478B - Power transmission line cotter pin target detection method based on multistage target detection - Google Patents
Power transmission line cotter pin target detection method based on multistage target detection Download PDFInfo
- Publication number
- CN115731478B CN115731478B CN202211484410.9A CN202211484410A CN115731478B CN 115731478 B CN115731478 B CN 115731478B CN 202211484410 A CN202211484410 A CN 202211484410A CN 115731478 B CN115731478 B CN 115731478B
- Authority
- CN
- China
- Prior art keywords
- cotter pin
- target
- target detection
- data
- level
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 94
- 230000005540 biological transmission Effects 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 14
- 238000012805 post-processing Methods 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims description 22
- 238000007781 pre-processing Methods 0.000 claims description 15
- 238000012795 verification Methods 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 7
- 230000000694 effects Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000007689 inspection Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 239000012212 insulator Substances 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000002372 labelling Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method for detecting a cotter pin target of a power transmission line based on multi-level target detection, which comprises the following steps of preparing a cotter pin data set for multi-level target detection; establishing a multi-stage target detector of the cotter pin, and obtaining marked objects of different layers and position information of the cotter pin through the multi-stage target detector; and (3) carrying out post-processing on the reasoning results of the multi-stage target detector, and judging whether the cotter pin exists and the existence state of the cotter pin by sorting and analyzing the target relations of different levels in the power line image and the types of the bolt small parts. The beneficial effects of the invention are as follows: the innovative multi-level target learning is provided, each level learns target object characteristics and cotter position characteristics of the level according to the marking information, and simultaneously the next level learns on the basis of characteristics generated by the previous level.
Description
Technical Field
The invention relates to a method for detecting a cotter pin target in a power transmission line, in particular to a method for detecting the cotter pin target of the power transmission line based on multi-stage target detection, and belongs to the technical field of power transmission line abnormality detection.
Background
The detection of the power transmission line cotter pin target needs to identify the characteristic points with tiny images, and the status of the cotter pin in the abnormal detection of the power transmission line contains important key information. Such as a cotter pin missing, may cause a large-area power failure due to disconnection of the transmission line, and even cause casualties. Therefore, identification of cotter pins in a power transmission line is of great importance.
The current cotter pin detection method mainly has the following defects:
(1) The cotter pin target is too small, and only 10-20 pixels or even a few pixels are in the image, so that false detection is easy to occur or existence of the cotter pin cannot be detected during algorithm processing, and therefore, the recognition accuracy is low, and the cotter pin recall rate is low;
(2) The detection data set is single in labeling information, only target labels needing to be identified are needed, for example, in a cotter pin target detection task, the data set only labels cotter pin information, and other auxiliary or multi-level labeling information is not available, so that an algorithm cannot learn by utilizing other effective information.
Disclosure of Invention
The invention aims to solve at least one technical problem, and provides a method for detecting the targets of cotter pins of a power transmission line based on multi-stage target detection, which effectively improves the recall rate and the accuracy rate of the target detection of cotter pins in the power transmission line
The invention realizes the above purpose through the following technical scheme: a method for detecting a power transmission line cotter pin target based on multistage target detection comprises the following steps of
Step one, preparing a cottage pin data set for multi-level target detection;
step two, establishing a multi-stage target detector of the cotter pin, training the multi-stage target detector with target detection learning capacity based on a cotter pin data set, and obtaining marked objects of different layers and obtaining position information of the cotter pin through the multi-stage target detector;
and thirdly, carrying out post-processing on the reasoning results of the multi-stage target detector, and judging whether the cotter pin exists or not and the existence state of the cotter pin by sorting and analyzing the target relations of different layers in the power line image and the types of the bolt small parts.
As still further aspects of the invention: in the first step, preparing a cotter pin data set for multi-level target detection includes: acquiring multi-level target detection cottage pin data; marking the multi-level target detection cotter pin data; and preprocessing the multi-level target detection data of the power transmission line.
As still further aspects of the invention: and acquiring the data of the multi-stage target detection cotter pin and the data of the inspection image of the unmanned aerial vehicle of the power transmission line including the cotter pin.
As still further aspects of the invention: the multi-level target detection cotter pin data marking refers to marking a target frame for collecting cotter pin images and a cotter pin center point, and marking in layers from small to large according to a target level:
first layer target annotation type: the cotter pin is positioned on an object;
second layer target annotation type: including the object on which the cotter pin is located, and so on.
As still further aspects of the invention: the transmission line multi-level target detection data preprocessing refers to preprocessing of a data set with multi-level target tags, wherein the multi-level targets comprise, but are not limited to, hanging points, wire clamps, grading rings, insulators, jumpers, bolts or cotter pins.
As still further aspects of the invention: the inclusion relationship and the peer relationship exist between different peer targets;
marking the layers of the hierarchy containing the relation layer by layer according to the size of the object from large to small;
if the cotter pin exists in the hierarchy of the peer relationship, the position of the cotter pin is taken as a center coordinate; if no cotter pin is present, the position coordinates are the origin.
As still further aspects of the invention: in the second step, the multi-stage object detector for establishing the cotter pin comprises: constructing an inference module of the multi-level target detector of the power transmission line; building a model structure; the model trains the inference modules of the multi-stage object detector.
As still further aspects of the invention: the construction of the inference module of the multi-level target detector of the power transmission line is based on preprocessing of multi-level target detection data of the power transmission line, hierarchical classification preprocessing is carried out on marked data, the marked data of a first layer of targets are used as training data of a first layer of inferators, and the like, and each layer of inferators has corresponding training data.
As still further aspects of the invention: each model training multi-stage object detector reasoning module is an object detection model with independent object detection function, and proper object detection models are selected according to reasoning tasks of different levels.
As still further aspects of the invention: in the third step, the post-processing of the reasoning result of the multi-stage target detector comprises the effect verification of the reasoning module of the multi-stage target detector and the post-processing of the reasoning data;
the reasoning module effect verification of the multi-stage target detector comprises dividing multi-stage target detection opening pin data into a training set, a verification set and a test set, wherein the test set is used for verifying the accuracy of multi-stage target detection;
the post-processing of the reasoning data comprises the following steps:
1) Using a multi-stage target detector to obtain each component type from the hanging point type of the power transmission line to the hanging point;
2) Carrying out layer-by-layer target detection on bolts in each component;
3) The cotter pin in the bolt is detected to obtain a state of existence of the cotter pin.
The beneficial effects of the invention are as follows: the detection of the cotter pin is split into multi-level tasks through learning marking information of different levels, the specific positions of the targets are identified from the large targets to the small targets, and meanwhile, the identification information of different levels is fused, so that the learning of multi-level information is realized, and the cotter pin is not only learned;
the innovation provides multi-level target learning, wherein each level learns target object characteristics and cotter position characteristics of the level according to marking information. Meanwhile, the next level learns on the basis of the characteristics generated by the previous level, the target information of each level in the marking information is fully utilized in the mode, and the detection result of the cotter pin is obtained through comprehensive learning, so that the recognition accuracy can be improved compared with other information only learning the characteristics of the cotter pin target.
Drawings
FIG. 1 is a schematic flow chart of the present invention;
FIG. 2 is a schematic diagram of an exemplary cotter pin test of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
As shown in fig. 1, a method for detecting a cotter pin target of a power transmission line based on multi-stage target detection includes the following steps:
first: a cotter pin dataset of a multi-level target detection is prepared.
The cotter pin dataset for multi-level target detection includes: acquiring multi-level target detection cottage pin data; marking the multi-level target detection cotter pin data; and preprocessing the multi-level target detection data of the power transmission line.
The multi-level target detection cotter pin data acquisition comprises power transmission line unmanned aerial vehicle inspection image data of the cotter pin, the acquired image data can be used for seeing an existing image of a tiny object, and the image can be clearly visible or can be a blurred image for seeing the outline of the tiny object.
The multi-level target detection cotter pin data labeling refers to labeling a target frame for collecting cotter pin images and cotter pin center points, and labeling in layers from small to large according to target levels:
first layer target annotation type: the cotter pin is positioned on an object;
second layer target annotation type: including the object on which the cotter pin is located, and so on.
The transmission line multi-level target detection data preprocessing refers to preprocessing of a data set with multi-level target tags, wherein the multi-level targets comprise, but are not limited to, hanging points, wire clamps, grading rings, insulators, jumpers, bolts or cotter pins.
The inclusion relationship and the peer relationship exist between different peer targets;
marking the layers of the hierarchy containing the relation layer by layer according to the size of the object from large to small;
if the cotter pin exists in the hierarchy of the peer relationship, the position of the cotter pin is taken as a center coordinate; if no cotter pin is present, the position coordinates are the origin.
Two ways are used for marking the target to be identified at the same time.
Second,: and establishing a multi-stage target detector of the cotter pin, training the multi-stage target detector with target detection learning capability based on the cotter pin data set, and obtaining marked objects of different layers and obtaining position information of the cotter pin through the multi-stage target detector.
The multi-stage object detector for establishing cotter pins includes: constructing an inference module of the multi-level target detector of the power transmission line; building a model structure; the model trains the inference modules of the multi-stage object detector.
The construction of the reasoning module of the multi-level target detector of the power transmission line is based on preprocessing of multi-level target detection data of the power transmission line, hierarchical classification preprocessing is carried out on marked data, the marked data of a first layer of targets are used as training data of a first layer of reasoners, and the like, and the reasoners of each hierarchy have corresponding training data.
The method for constructing the reasoning modules of the multi-stage target detector of the power transmission line is mainly used for solving the problem that each reasoning module of the multi-stage target detector is a target detection model with an independent target detection function, the target detection model selects a proper target detection model according to different reasoning tasks of different levels, for example, the model improvement is performed on the detection of a first-level large target by selecting yolo with high reasoning speed, the target detection is performed, and the target detection of a small target is performed by adopting a two-section target detection model.
The model structure is built and is mainly used for obtaining the reasoning result of the multi-stage target detector, each stage of target detection module can reasoning and identify targets under the level, and thus targets of different levels can be identified by multi-stage target detection of one power transmission line picture. And establishing serial connection and parallel connection of different reasoning modules according to the marked different hierarchical relationships.
The reasoning module method for training the multi-stage target detector is mainly used for each module to have independent loss function and is used for training the target detection module model. And updating parameters of the target detection module by using an optimizer, and realizing optimization of the model through continuous iterative computation.
The reasoning module of the model training multi-stage target detector is mainly used for training a reasoning module method of the multi-stage target detector, is used for each module to have an independent loss function, is used for training the model of the target detection module, uses an optimizer to update parameters of the target detection module, and realizes optimization of the model through continuous iterative computation. Each model training multi-stage object detector reasoning module is an object detection model with independent object detection function, and proper object detection models are selected according to reasoning tasks of different levels.
Third,: and (3) carrying out post-processing on the reasoning results of the multi-stage target detector, and judging whether the cotter pin exists and the existence state of the cotter pin by sorting and analyzing the target relations of different levels in the power line image and the types of the bolt small parts.
The reasoning result post-processing of the multi-stage target detector comprises the effect verification of a reasoning module of the multi-stage target detector and the post-processing of reasoning data;
the reasoning module effect verification of the multi-stage target detector comprises dividing multi-stage target detection opening pin data into a training set, a verification set and a test set, wherein the test set is used for verifying the accuracy of multi-stage target detection;
the post-processing of the reasoning data comprises the following steps:
1) Using a multi-stage target detector to obtain each component type from the hanging point type of the power transmission line to the hanging point;
2) Carrying out layer-by-layer target detection on bolts in each component;
3) The cotter pin in the bolt is detected to obtain a state of existence of the cotter pin.
Example two
As shown in fig. 2, a method for detecting a small part image target of a power transmission line based on multi-level target detection is provided, wherein the multi-level target detection is used for detecting each component type from a hanging point type to a hanging point of the power transmission line, then the bolts in each component are subjected to layer-by-layer target detection, and finally cotter pins in the bolts are detected. By this method, it is possible to obtain an identification of the existence state of the cotter pin.
The method comprises the following steps:
firstly, performing primary target detection and inputting: inspection images, output: a ground wire hanging point characteristic diagram of cotter pin position characteristics;
step two, performing secondary target detection and inputting: cotter pin coordinate feature vector and image feature map, output: a clip signature of cotter pin position features;
third, three-level target detection is carried out, and input is carried out: cotter pin coordinate feature vector and image feature map, output: a bolt signature of cotter pin position features;
outputting a target detection result: cotter pin detection results.
Working principle: preparing a cotter pin data set for multi-level target detection, then establishing a cotter pin multi-level target detector, training the multi-level target detector by using the prepared cotter pin data set for multi-level target detection, and finally post-processing the reasoning result of the multi-level target detector.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present disclosure describes embodiments, not every embodiment is provided with a separate embodiment, and that this description is provided for clarity only, and that the disclosure is not limited to the embodiments described in detail below, and that the embodiments described in the examples may be combined as appropriate to form other embodiments that will be apparent to those skilled in the art.
Claims (3)
1. A power transmission line cotter pin target detection method based on multi-stage target detection is characterized by comprising the following steps of: comprises the following steps
Step one, preparing a cottage pin data set for multi-level target detection;
in the first step, preparing a cotter pin data set for multi-level target detection includes: acquiring multi-level target detection cottage pin data; marking the multi-level target detection cotter pin data; preprocessing multi-level target detection data of the power transmission line;
the multi-level target detection cotter pin data marking refers to marking a target frame for collecting cotter pin images and a cotter pin center point, and marking in layers from small to large according to a target level:
first layer target annotation type: the cotter pin is positioned on an object;
second layer target annotation type: on the object including the object on which the cotter pin is located, and so on;
the power transmission line multistage target detection data preprocessing refers to preprocessing of a data set with multistage target labels, wherein the multistage targets comprise hanging points, wire clamps, equalizing rings, insulators, jumpers, bolts or cotter pins;
the inclusion relationship and the peer relationship exist between different peer targets;
marking the layers of the hierarchy containing the relation layer by layer according to the size of the object from large to small;
if the cotter pin exists in the hierarchy of the peer relationship, the position of the cotter pin is taken as a center coordinate; if the cotter pin does not exist, the position coordinate is the origin;
step two, establishing a multi-stage target detector of the cotter pin, training the multi-stage target detector with target detection learning capacity based on a cotter pin data set, and obtaining marked objects of different layers and obtaining position information of the cotter pin through the multi-stage target detector;
in the second step, the multi-stage object detector for establishing the cotter pin comprises: constructing an inference module of the multi-level target detector of the power transmission line; building a model structure; model training the reasoning module of the multi-stage target detector;
the construction of the reasoning module of the multi-level target detector of the power transmission line is based on preprocessing of multi-level target detection data of the power transmission line, hierarchical classification preprocessing is carried out on marked data, the marked data of a first layer of targets are used as training data of a first layer of reasoners, and the reasoning modules of each layer have corresponding training data;
step three, post-processing of reasoning results of the multi-stage target detector, and judging whether the cotter pin exists and the existence state of the cotter pin by sorting and analyzing target relations of different levels in the power line image and the types of the bolt small parts;
in the third step, the post-processing of the reasoning result of the multi-stage target detector comprises the effect verification of the reasoning module of the multi-stage target detector and the post-processing of the reasoning data;
the reasoning module effect verification of the multi-stage target detector comprises dividing multi-stage target detection opening pin data into a training set, a verification set and a test set, wherein the test set is used for verifying the accuracy of multi-stage target detection;
the post-processing of the reasoning data comprises the following steps:
1) Using a multi-stage target detector to obtain each component type from the hanging point type of the power transmission line to the hanging point;
2) Carrying out layer-by-layer target detection on bolts in each component;
3) The cotter pin in the bolt is detected to obtain a state of existence of the cotter pin.
2. The transmission line cotter pin target detecting method according to claim 1, characterized in that: and acquiring the data of the multi-stage target detection cotter pin and the data of the inspection image of the unmanned aerial vehicle of the power transmission line including the cotter pin.
3. The transmission line cotter pin target detecting method according to claim 1, characterized in that: each model training multi-stage object detector reasoning module is an object detection model with independent object detection function, and proper object detection models are selected according to reasoning tasks of different levels.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211484410.9A CN115731478B (en) | 2022-11-24 | 2022-11-24 | Power transmission line cotter pin target detection method based on multistage target detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211484410.9A CN115731478B (en) | 2022-11-24 | 2022-11-24 | Power transmission line cotter pin target detection method based on multistage target detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115731478A CN115731478A (en) | 2023-03-03 |
| CN115731478B true CN115731478B (en) | 2023-12-22 |
Family
ID=85298077
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211484410.9A Active CN115731478B (en) | 2022-11-24 | 2022-11-24 | Power transmission line cotter pin target detection method based on multistage target detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115731478B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110688925A (en) * | 2019-09-19 | 2020-01-14 | 国网山东省电力公司电力科学研究院 | Cascade target identification method and system based on deep learning |
| CN111815623A (en) * | 2020-07-28 | 2020-10-23 | 南方电网数字电网研究院有限公司 | Power transmission line cotter pin missing identification method |
| CN114170144A (en) * | 2021-11-11 | 2022-03-11 | 国网福建省电力有限公司漳州供电公司 | Power transmission line pin defect detection method, equipment and medium |
| WO2022111219A1 (en) * | 2020-11-30 | 2022-06-02 | 华南理工大学 | Domain adaptation device operation and maintenance system and method |
| CN114998576A (en) * | 2022-08-08 | 2022-09-02 | 广东电网有限责任公司佛山供电局 | Method, device, equipment and medium for detecting loss of cotter pin of power transmission line |
| CN115619763A (en) * | 2022-10-31 | 2023-01-17 | 国网湖北省电力有限公司超高压公司 | Power transmission line small part image target detection method based on multistage reasoning detection |
-
2022
- 2022-11-24 CN CN202211484410.9A patent/CN115731478B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110688925A (en) * | 2019-09-19 | 2020-01-14 | 国网山东省电力公司电力科学研究院 | Cascade target identification method and system based on deep learning |
| CN111815623A (en) * | 2020-07-28 | 2020-10-23 | 南方电网数字电网研究院有限公司 | Power transmission line cotter pin missing identification method |
| WO2022111219A1 (en) * | 2020-11-30 | 2022-06-02 | 华南理工大学 | Domain adaptation device operation and maintenance system and method |
| CN114170144A (en) * | 2021-11-11 | 2022-03-11 | 国网福建省电力有限公司漳州供电公司 | Power transmission line pin defect detection method, equipment and medium |
| CN114998576A (en) * | 2022-08-08 | 2022-09-02 | 广东电网有限责任公司佛山供电局 | Method, device, equipment and medium for detecting loss of cotter pin of power transmission line |
| CN115619763A (en) * | 2022-10-31 | 2023-01-17 | 国网湖北省电力有限公司超高压公司 | Power transmission line small part image target detection method based on multistage reasoning detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115731478A (en) | 2023-03-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111339882B (en) | Power transmission line hidden danger detection method based on example segmentation | |
| CN112183788B (en) | Domain adaptive equipment operation detection system and method | |
| CN110569841B (en) | Contact gateway key component target detection method based on convolutional neural network | |
| CN114627360A (en) | Substation equipment defect identification method based on cascade detection model | |
| CN111008961B (en) | Transmission line equipment defect detection method and system, equipment and medium thereof | |
| CN110992349A (en) | Underground pipeline abnormity automatic positioning and identification method based on deep learning | |
| CN107004266A (en) | The method for detecting defect on surface of tyre | |
| CN114331949A (en) | Image data processing method, computer equipment and readable storage medium | |
| CN111598855B (en) | 2C equipment high-speed rail contact net dropper defect detection method based on deep learning and transfer learning | |
| US20200034626A1 (en) | Method for recognizing an object of a mobile unit | |
| CN117541534A (en) | Power transmission line inspection method based on unmanned plane and CNN-BiLSTM model | |
| CN116823793A (en) | Device defect detection method, device, electronic device and readable storage medium | |
| Lam et al. | Real-time traffic status detection from on-line images using generic object detection system with deep learning | |
| CN115546223A (en) | Method and system for detecting loss of fastening bolt of equipment under train | |
| CN116486240A (en) | Application of image recognition algorithm in intelligent inspection method of unmanned aerial vehicle of power transmission line | |
| CN117252815A (en) | Industrial part defect detection method, system, equipment and storage medium based on 2D-3D multi-mode image | |
| Manninen et al. | Multi-stage deep learning networks for automated assessment of electricity transmission infrastructure using fly-by images | |
| CN112837281B (en) | Pin defect identification method, device and equipment based on cascade convolution neural network | |
| CN115731478B (en) | Power transmission line cotter pin target detection method based on multistage target detection | |
| CN115619763A (en) | Power transmission line small part image target detection method based on multistage reasoning detection | |
| CN117115727A (en) | Transformer substation defect judging method and system | |
| CN115049836B (en) | Image segmentation method, device, equipment and storage medium | |
| CN116309407A (en) | Method for detecting abnormal state of railway contact net bolt | |
| US20230084761A1 (en) | Automated identification of training data candidates for perception systems | |
| CN114638989A (en) | Fault classification visualization method based on target detection and fine-grained identification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |